Simulation and Characterization of Cathode Reactions in Solid Oxide Fuel Cells

Abstract

In this study, we have developed a dense La0.85Sr0.15MnO3-δ (LSM) Ce0.9Gd0.1O1.95 (GDC) composite electrode system for studying the surface modification of cathodes. The LSM and GDC grains in the composite were well defined and distinguished using energy dispersive x-ray (EDX) analysis. The specific three-phase boundary (TPB) length per unit electrode surface area was systematically controlled by adjusting the LSM to GDC volume ratio of the composite from 40% up to 70%. The TPB length for each tested sample was determined through stereological techniques and used to correlate the cell performance and degradation with the specific TPB length per unit surface area. An overlapping spheres percolation model was developed to estimate the activity of the TPB lines on the surface of the dense composite electrodes developed. The model suggested that the majority of the TPB lines would be active and the length of those lines maximized if the volume percent of the electrolyte material was kept in the range of 47 57%. Additionally, other insights into the processing conditions to maximize the amount of active TPB length were garnered from both the stereology calculations and the percolation simulations. Steady-state current voltage measurements as well as electrochemical impedance measurements on numerous samples under various environmental conditions were completed. The apparent activation energy for the reduction reaction was found to lie somewhere between 31 kJ/mol and 41 kJ/mol depending upon the experimental conditions. The exchange current density was found to vary with the partial pressure of oxygen differently over two separate regions. At relatively low partial pressures, i0 had an approximately dependence and at relatively high partial pressures, i0 had an approximately dependence. This led to the conclusion that a change in the rate limiting step occurs over this range. A method for deriving the electrochemical properties from proposed reaction mechanisms was also presented. State-space modeling was used as it is a robust approach to addressing these particular types of problems due to its relative ease of implementation and ability to efficiently handle large systems of differential algebraic equations. This method combined theoretical development with experimental results obtained previously to predict the electrochemical performance data. The simulations agreed well the experimental data and allowed for testing of operating conditions not easily reproducible in the lab (e.g. precise control and differentiation of low oxygen partial pressures).